ORIGINAL PAPER

Performance of biocomposites from surface modifiedregenerated cellulose fibers and lactic acid thermosetbioresin

Sunil Kumar Ramamoorthy . Fatimat Bakare .

Rene Herrmann . Mikael Skrifvars

Received: 4 February 2015 / Accepted: 29 April 2015

� Springer Science+Business Media Dordrecht 2015

Abstract The effect of surface treatments, silane and

alkali, on regenerated cellulose fibers was studied by

using the treated fibers as reinforcement in lactic acid

thermoset bioresin. The surface treatments were

performed to improve the physico–chemical interac-

tions at the fiber–matrix interface. Tensile, flexural and

impact tests were used as indicator of the improvement

of the interfacial strength. Furthermore, thermal con-

ductivity, viscoelasticity measurements as well as

microscopy images were made to characterize the fiber

surface treatments and the effect on adhesion to the

matrix. The results showed that silane treatment

improved the mechanical properties of the composites

as the silane molecule acts as link between the

cellulose fiber and the resin (the fiber bonds with

siloxane bridge while the resin bonds with organofunc-

tional group of the bi-functional silane molecule)

which gives molecular continuity in the interphase of

the composite. Porosity volume decreased significant-

ly on silane treatment due to improved interface and

interlocking between fiber and matrix. Decrease in

water absorption and increase in contact angle con-

firmed the change in the hydrophilicity of the com-

posites. The storage modulus increased when the

reinforcements were treated with silane whereas the

damping intensity decreased for the same composites

indicating a better adhesion between fiber and matrix

on silane treatment. Thermogravimetric analysis indi-

cated that the thermal stability of the reinforcement

altered after treatments. The resin curing was followed

using differential scanning calorimetry and the neces-

sity for post-curing was recommended. Finite element

analysis was used to predict the thermal behavior of

the composites and a non-destructive resonance

analysis was performed to ratify the modulus obtained

from tensile testing. The changes were also seen on

composites reinforced with alkali treated fiber. Mi-

croscopy images confirmed the good adhesion be-

tween the silane treated fibers and the resin at the

interface.

Keywords Surface modification � Cellulose fiber �Mechanical properties � Thermal conductivity � Finiteelement analysis � Resonance analysis

Introduction

Natural fiber reinforced composites have shown

significant importance in the composite industry and

the good reinforcing potential of these fibers has led to

extensive use of these composites in automotive

applications (Jawaid and Abdul Khalil 2011; La

S. K. Ramamoorthy � F. Bakare � M. Skrifvars (&)

Swedish Centre for Resource Recovery, University of

Boras, 501 90 Boras, Sweden

e-mail: Mikael.Skrifvars@hb.se

R. Herrmann

Department of Energy and Materials Technology, Arcada

University of Applied Science, 00560 Helsinki, Finland

123

Cellulose

DOI 10.1007/s10570-015-0643-x

Mantia and Morreale 2011; Faruk et al. 2012; Bledzki

and Gassan 1999; Koronis et al. 2013; Puglia et al.

2005). The use of cellulosic fibers obtained from bast,

leaf, seed, wood, straw and grass as reinforcements in

plastics has increased in recent decades (Faruk et al.

2012; Ramamoorthy et al. 2015). Natural fibers have

good physical and mechanical properties, which make

them a right choice as low cost reinforcement (Faruk

et al. 2012; Bledzki and Gassan 1999; Ramamoorthy

et al. 2015). However, the hydrophilicity due to the

cellulosic hydroxyl groups and surface unevenness

make natural fibers less preferred over synthetic fibers

such as glass fibers in more demanding composite

applications. The mechanical processing from field to

fiber includes several stages (growing, field retting,

scotching, hackling, carding and spinning), this makes

the processing both costly, and also prone to quality

fluctuations (Jawaid and Abdul Khalil 2011; Faruk

et al. 2012). Regenerated or partly man-made cellulose

fibers do not have these drawbacks.

Researchers have used regenerated cellulosic fibers

as an alternate to the natural fibers in fiber reinforced

composites (FRC); as the primary constituents’ of

both the fibers remains the same. Regenerated cellu-

lose fibers are chemically pure, their surface structure

is even, and the fiber properties can be reproduced

easily (Woodings 2001; Carrillo et al. 2010; Adusu-

mali et al. 2006; Fink et al. 2014). It is also found that

these fibers have good potential to be used as

reinforcement in FRC (Carrillo et al. 2010; Adusumali

et al. 2006; Ganster and Fink 2006; Jaturapiree et al.

2006). High surface evenness and even quality of

these fibers makes it possible to get consistent results

which are not possible in natural fiber (Woodings

2001). These fibers are unique, as they possess the

beneficial properties of both natural and synthetic

fibers (Johnson et al. 2008). They have low density,

which give low weight composites, their environmen-

tal characteristics are like as for natural fibers, and

their performance is stable like as for synthetic fibers

(Johnson et al. 2008).

Our previous studies show the great potential of

regenerated cellulose fibers as reinforcements in FRC,

and these could be a good alternative for annual plant

fibers (Ramamoorthy et al. 2012, 2013, 2014). The

mechanical properties of composites reinforced with

regenerated cellulose are better than for several natural

fiber composites (Carrillo et al. 2010; Ramamoorthy

et al. 2012, 2013; Saevey et al. 2001).

In recent years, the synthesis of biothermosets has

been reported. The matrix is synthesized from renew-

able constituents such as soybean or linseed oil, or

from lactic acid (Seniha Guner et al. 2006; Akesson

et al. 2010). The properties of synthesized matrixes

from renewable resources have been explored by

researchers and proven to be good alternative to

petroleum based matrixes in natural fiber reinforced

composites (Akesson et al. 2011; Adekunle et al.

2011). Reinforcements and matrix resins made from

renewable resources have been claimed to have

environmental and economic benefits such as low

pollutant emissions, low greenhouse gas emissions

and low cost (La Mantia and Morreale 2011).

Researchers are therefore aiming to produce green

composites for several applications from renewable

reinforcement and matrix (La Mantia and Morreale

2011).

However, there are several issues involved when

using polar hydrophilic fibers with non-polar hy-

drophobic matrixes in the production of biocompos-

ites, and the most important issue is the interface

between fiber and matrix (John and Anandjiwala

2008). Cellulosic fibers have polar OH groups in each

glucan unit and can readily absorb moisture or water

while the thermoset matrixes derived from lactic

acid are nearly hydrophobic which causes poor

fiber–matrix interface due to poor fiber wetting

(Caulfield et al. 1999). The mechanical properties of

the composites are directly affected by the interfacial

adhesion as the load transfer takes place by shear

stresses at the interface (Caulfield et al. 1999). Poor

fiber–matrix adhesion in the interface results in

improper transfer of load and thus reducing the full

capabilities of the composite (Caulfield et al. 1999).

Therefore, the natural fibers are often modified,

physically or chemically, to attain good interface,

which improves mechanical performance of the

composites (Mohanty et al. 2001; Bledzki et al.

1996).

Chemical treatments such as silane and alkali are

common and straight forward methods to modify the

surface of the natural fibers which has improved the

properties (Mohanty et al. 2001). Such treatments also

roughen the surface of the fiber, which improves the

mechanical bonding and thus stress transfer at inter-

face is appreciable (Mohanty et al. 2001). Alkali

treatment disrupts the intermolecular hydrogen bond-

ing in the cellulose fiber bundle and removes certain

Cellulose

123

parts (lignin, hemicellulose, wax and oil) of the fibers,

which alter the fiber surface (Mohanty et al. 2001).

Silane treatment acts as a coupling agent that may

reduce the number of hydroxyl groups in the fiber–

matrix interface by reaction with the silica molecules

(Ramamoorthy et al. 2014). These changes in the

fibers’ surface could improve the mechanical proper-

ties of the fibers due to change in structural changes

(John and Anandjiwala 2008). The fiber–matrix inter-

face is also affected by the change in the fibers’

characteristics as the physico–chemical interactions

between fiber and matrix will be altered (Mohanty

et al. 2001).

Our previous studies show that the mentioned

treatments improve the surface roughness and the

performance of the fibers for composite applications

(Ramamoorthy et al. 2014); and will result in a

good reinforcement for of the lactic acid based

biomatrix (Bakare et al. 2014). A untreated regen-

erated cellulose reinforced soybean-based thermoset

composites show even a better performance than

some natural fiber composites (Faruk et al. 2012;

Carrillo et al. 2010; Ramamoorthy et al. 2012,

2013). This is mainly attributed to superior fiber

properties. Microscopic images indicated that these

composites were porous; and the unevenness in

matrix impregnation resulted in delamination (Ra-

mamoorthy et al. 2013). The absence of chemical

treatment leads to poor interface between fiber and

matrix (Caulfield et al. 1999; Mohanty et al. 2001;

Bledzki et al. 1996). In recent study regenerated

cellulose fibers were treated by silane and alkali and

the treatment conditions such as treatment time,

treatment temperature and concentration of the

treatment solution were studied. Testing of the

treated fibers revealed better properties of some of

the treated fibers than untreated fibers. Harsh

treatment conditions lead to fibrillation of fibers,

which resulted in poor mechanical performance

(Ramamoorthy et al. 2014). Biomatrix from lactic

acid has been synthesized and characterized in our

prior study which revealed the good potential of the

matrix in composite industry (Bakare et al. 2014).

In this paper, composites are produced from surface

treated regenerated cellulose fibers and a lactic acid

thermoset resin. The composite performance is

evaluated through mechanical, thermal, viscoelastic

and morphological properties.

Experimental

Materials

Viscose nonwoven fabrics with 60 g/m2 surface

weight and made from 1.7 dtex linear density regen-

erated cellulose staple fibers were supplied by Suomi-

nen Nonwovens Ltd, Finland. Reagent grade sodium

hydroxide pellets, L-lactic acid (88–92 %) and glyc-

erol (99.5 %) were obtained from Scharlau, Spain.

APTES, 3-aminopropyl-triethoxysilane (99 %),

toluene (99.9 %), absolute ethanol (C99.8 %), ethy-

lene glycol (99.8 %), formamide (C99.5 %) and

hexadecane ([99 %) were supplied by Sigma Aldrich,

USA. Dibenzoyl peroxide initiator ([98 %) and

hydroquinone (99 %) were supplied by Kebo Lab,

Sweden and Acros Organics respectively. Methane-

sulfonic acid (C98 %) and methacrylic anhydride

were supplied by Alfa Aesar.

Treatment

Alkali treatment

Regenerated cellulose fabrics were pre-dried at

105 �C before immersing in three different concen-

trations (6, 8 and 10 wt%) of sodium hydroxide

(NaOH) solutions for 30 min at room temperature.

After the treatment, fibers were washed thoroughly

with distilled water until neutrality. The pH was

checked periodically using litmus paper. Then the

fibers were dried in room temperature followed by

oven drying for 3 h at 105 �C. Fiber behavior on alkalitreatment was studied in previous research (Ra-

mamoorthy et al. 2014; Jaturapiree et al. 2006;

Okubayashi and Bechtold 2005).

Silane treatment

Pre-dried regenerated cellulose fabrics were immersed

in three different silane concentration solutions (6, 8

and 10 wt%) at room temperature for 30 min. APTES

(3-aminopropyl-triethoxysilane) was used as silane

coupling agent and was added to ethanol–water

mixture (8:2 volume ratio) to treat the regenerated

cellulose fabrics. Fibers were washed thoroughly with

distilled water after treatment and pH was checked for

neutrality. Then the fibers were dried in room

Cellulose

123

temperature followed by oven drying for 3 h at 105 �C(Fig. 1).

In our previous research, the spectral changes were

followed by Fourier transform infrared spectroscopy

(FTIR) and Raman spectroscopy, which confirmed the

changes in the surface chemistry of the fibers on silane

and alkali treatments (Ramamoorthy et al. 2014).

Detailed investigation on treated fibers was presented

in our preceding study (Ramamoorthy et al. 2014).

Resin synthesis

The synthesis was done according to the procedure

discussed thoroughly in our earlier work (Bakare et al.

2014). Lactic acid and glycerol were reacted in a direct

condensation; the product was then end-functionalized

with methacrylic anhydride. The reaction was done in

a 3-neck flask, equipped with a Dean-Stark distilling

head and a condenser. In the first step, 0.12 mol of

glycerol was added to 1.08 mol of lactic acid diluted in

50 g of toluene. Methanesulphonic acid, 0.1 wt%, was

used as catalyst. The condensation reaction took place

for totally about 5 h, first 2 h at 145 �C, followed by

2 h at 165 �C and 1 h at 195 �C. In the second step, theproduct was cooled to 110 �C and end-functionalized

with 0.396 mol of methacrylic anhydride. The end-

functionalization proceeded for about 4 h. Both reac-

tion steps were done under nitrogen atmosphere. The

resin was obtained by removing the residual

methacrylic acid that had formed after the second

step together with toluene using rotary evaporator

distillation at a temperature of 60 �C and a pressure of

13 mbar.

Figure 2 shows the two reactions, condensation and

functionalization, during resin synthesis. The viscosity

of the resin was 1.09 Pa s. The bio content was

calculated according to ASTM D6866 using the

equations:

Bio content ¼ bio carbon content

total carbon content� 100 ð1Þ

Bio content ¼ Gm þ 3n Lm �Wmð Þ½ �Gm þ 3 MAm �Mmð Þ þ n Lm �Wmð Þ½ �� 100

ð2Þ

where Gm is the weight of glycerol, n is the chain

length, Lm is the weight of lactic acid, Wm is the

weight of the water condensed, MAm is the weight of

methacrylic anhydride and Mm is the weight of the

H3C

CH2

O

Si

H2C

CH2

H2C

NH2

O OH2C CH2

H3C CH3

Fig. 1 Structural formula of 3-aminopropyl-triethoxysilane

(C9H23NO3Si)

Fig. 2 Resin synthesis

reaction. a Condensation

reaction, b methacrylic

functionalization

Cellulose

123

methacrylic acid formed (Bakare et al. 2014). Based

on this equation, it was found that the resin had 78 %

bio content.

The 13C-NMR results presented in our previous

paper showed that 90.2 % lactic acid reacted with

glycerol whereas 8.9 % reacted into lactic acid

oligomers and 0.9 % formed a lactide. The chain

length of the polymer and the lactic acid oligomers

were 4 and the 2.2 respectively. Functionalization with

methacrylic anhydride in the second step was verified

using Fourier transform infrared (FTIR).

Composite preparation

The regenerated cellulose fabrics, untreated and

treated, were cut to 20 9 20 cm2 dimension. Resin

viscosity has a major role in fiber impregnation and the

viscosity was therefore reduced by heating the resin in

an oven at 60 �C for 5 min. Dibenzoyl peroxide

(2 wt%) was used as initiator and was mixed well with

the heated resin to get a homogeneous mixing. The

fiber–matrix weight ratio of about 1:1 was chosen due

to good resin impregnation. The impregnated fabrics

were then stacked between two steel plates and

pressed in a manual hot press. The curing was done

at a temperature of 150 �C and at a pressure of 40 bar.

The hot press used for the compression molding was

supplied by Rondol Technology, Staffordshire, UK.

Specimen preparation

Specimens were cut from the composite laminates

using laser-cutting technology, GCC LaserPro Spirit

(Taiwan), for below mentioned tests according to ISO

standards. The specimens were cut from the centers of

the composite laminates, leaving the edges to mini-

mize edge effects.

Characterisation

Tensile test

Tensile tests were carried out according to ISO 527

using a Tinius Olsen H10KT universal testingmachine

and QMat software. The dumb bell shaped tensile test

specimens were clamped and pulled apart; using a 10

kN load cell and a mechanical extensiometer. The

overall length of the specimen was 150 mm which

includes 60 mm parallel sided portion. The parallel

sided portions’ width was 10 mm and the width at the

ends was 20 mm. The gauge length was 50 mm and

the rate of loading was 10 mm/min. The tensile

properties such as tensile strength, E-modulus and %

elongation were studied. Ten specimens were tested

for each batch and the average is noted.

Impact test

The Charpy impact test was performed according to

ISO 179 using a digital Zwick test instrument. The flat

and un-notched specimens were tested flatwise. Ten

specimens were tested for each batch and the average

impact resistance is noted.

Flexural test

The three point flexural tests were carried out accord-

ing to ISO 14125 using Tinius Olsen H10KT universal

testing machine and QMat software. The flat speci-

mens, 80 9 15 mm2 (length 9 width), were placed

with support at two ends and the force was applied in

the middle; and this force was detected by load cell

(10 kN). The outer span was 64 mm and the rate of

loading was 10 mm/min. The flexural properties such

as flexural strength and modulus were investigated.

Five specimens were tested for each batch and the

average is noted.

Density and porosity measurements

Archimedes principle, buoyancy method, was used to

determine the densities of the composite specimens.

The specimens were covered with paraffin before

immersing in ethanol medium to avoid absorption.

The porosity was calculated using the method sug-

gested by Madsen et al. (2007). At least three

measurements were done for each sample and the

average is reported.

Dynamic mechanical thermal analysis (DMTA)

Dynamic mechanical thermal analysis was carried out

using Q series DMTA, TA Instrument, supplied by

Waters LLC, USA. The specimen dimension was

30 9 10 mm2 (length 9 width), and it is mounted in

dual cantilever clamp. The temperature range was

between 20 and 150 �C at a frequency of 1 Hz. The

Cellulose

123

storage modulus, loss modulus and tan d were studiedto evaluate the viscoelastic performance of the com-

posites. The measurements were done at least twice

for each sample.

Differential scanning calorimetry (DSC)

Differential scanning calorimetry analysis was carried

out using Q series DSC, TA Instrument, supplied by

Waters LLC, USA. Approximately 10 mg of compos-

ite sample sealed in aluminum pan was heated from 20

to 200 �C at a rate of 10 �C/min. The sample is cooled

to 20 �C and heated again to 200 �C. The analysis wasdone under nitrogen. The measurements were done at

least twice for each sample. DSC was also used to

measure specific heat capacity (Cp) of the fiber and the

cured resin. The instrument was calibrated with

standard sapphire material before the experimentation,

at the temperature range of 0 and 54 �C, at a heating

rate of 10 �C/min. Measurements were done twice

with approximately 13 mg sample in nitrogen

atmosphere.

Thermogravimetric analysis (TGA)

Thermogravimetric analysis was carried out using Q

series TGA, TA Instrument, supplied by Waters LLC,

USA. The sample weighing approximately 15 mg was

heated from 30 to 600 �C in platinum pan at a heating

rate of 10 �C/min under nitrogen atmosphere. The

measurements were done at least twice for each

sample and the average is reported.

Contact angle and surface energy

The contact angle measurement was done on optical

tensiometer from Attension, Theta Lite (Espoo, Fin-

land) equipped with appropriate software. A liquid

drop is placed on flat surface of the composite and the

contact angle is measured using camera by assessing

shape of the drop. The contact angle is directly

measured from the angle formed between the com-

posite sample and the tangent to the drop surface. Four

wetting agents such as water (WA), ethylene glycol

(EG), formamide (FO) and hexadecane (HD) were

used. The contact angles and the surface energies of

the composites reinforced with untreated and treated

fibers were measured. At least seven measurements

were done for each sample and the average is reported.

Water absorption

Gravimetric water absorption tests were carried out on

the composites (36 9 12 mm2), by immersing in

water for 10 days to determine the hydrolytic stability

of the composites. The specimens were dried and

weighed before placing in the distilled water at room

temperature and pressure. The amount of water

absorbed was measured everyday by taking out the

specimen and carefully drying the specimen surface

before weighing. The thickness of the specimens was

measured on a daily basis to study the stability. The

percentage water absorption (WA%) and thickness

increase were studied. At least three measurements

were done for each sample and the average is reported.

Microscopy

Morphological analysis was performed on cross-

section of tensile fractured specimens using optical

and scanning electron microscope (SEM). The optical

microscope was supplied by Nikon Instruments

(Japan) and the images were taken directly on cross-

section without any preparation. Scanning electron

microscope images were taken on AIS2100 (Seron

Technology, Korea) at 20 kV accelerated voltage after

the specimens were sputtered with a layer of gold.

Resonance analysis

Resonance frequency after bending the specimen was

obtained by a non-destructive method using a com-

posite cantilever beam. The apparatus for the analysis

was self-constructed by attaching a piezo electric

transducer to the beam which was connected to a

laptop’s sound card via 2.5 mm audio jack after

passing through high pass filter in series. The beam

was excited by bending, and the acoustic signals were

received using Audacity software. The signals were

converted to frequency and further analyzed using

Scilab 5.5.0 numerical computational software. This is

compared to the Young’s modulus obtained by the

mechanical testing.

The Young’s modulus is calculated using the

following equation,

Cellulose

123

x ¼ffiffiffiffi

k

m

r

¼ a2ffiffiffiffiffiffiffi

EI

kL4

r

ð3Þ

where E is the young’s modulus, I is the moment of

inertia of a beam, k is the linear mass density and L is

the length of the beam. This can be re-arranged to find

the Young’s modulus,

E ¼ kI

� �

A

a2

� �2

where A ¼ a2ffiffiffiffiffi

EI

k

r

ð4Þ

Using this equation, E ¼ kI

� �

Aa2� �2

; the modulus was

calculated; k was calculated by measuring the mass

and length of the beam, the I of the beam was

calculated using thickness and width width�thickness3

12

� �

;

A was the slope obtained from the Fig. 7c and a is

constant value, 1.875.

Code executed in Scilab 5.5.0 from audio signals

obtained from Audacity,

Thermo-physical test

Thermo-physical study was performed according to

ISO 22007 on transient plane source (TPS) 2500 hot

disko (Hot Disk AB, Gothenburg, Sweden) equipped

with suitable software at room temperature and

pressure. The specimen dimension was 50 9

10 mm2 and thickness was between 3 and 4 mm.

The heating power and the measurement time were

50 mW and 10 s. Two identical specimens were used

for each measurement; one above the sensor and

another underneath the sensor. The heat source is used

to heat the specimen and the date is collected from

both the specimens at the same time. The software

reported the average thermal properties of the two

specimens. Thermal conductivity, thermal diffusivity

and volumetric specific heat were measured simulta-

neously. At least three measurements were done for

each sample and the average is reported.

Thermal conductivity can also be calculated using

Bruggman’s model,

ke � kf� �

ðkmÞ13

km � kf� �

ðkeÞ13

¼ 1� vð Þ ð5Þ

where kf, and km are thermal conductivity of the fiber

and matrix respectively, ke is the efficient thermal

conductivity of the composite.

Finite element analysis

Finite element analysis software, COMSOL 4.3b, was

used to predict the temperature rise generated in the

composites during the heating experiments. The

specific heat capacity of the fiber and cured resin

was obtained from the DSC analysis and was used to

simulate the experimental method. The temperature

distribution for the composite during heating was

calculated by a Fourier heat conduction equation:

Cellulose

123

qCp

oT

ot¼ kx

o2T

ox2þ ky

o2T

oy2þ kz

o2T

oz2ð6Þ

where q is the density and Cp is the specific heat

capacity of the composite; and kx, ky and kz are

thermal conductivity coefficients in main axes. For the

nonwoven (unidirectional) composites, the thermal

conductivity in machine direction can be calculated by

the rule of mixtures

kc ¼ vf kf þ 1� vf� �

km ð7Þ

where kc, kf and km are thermal conductivity of the

composite, fiber and matrix; and Vf is fiber volume

fraction.

Results and discussion

The change in chemical composition and the crys-

tallinity after silane and alkali treatments on cellulose

fibers were reported in a former study using FTIR and

Raman spectroscopy (Ramamoorthy et al. 2014).

Mechanical (tensile strength, modulus and elongation)

and thermal (heat capacity and stability) properties of

the treated fibers were investigated. Further, weight

loss, swelling andmoisture absorption were examined.

The hydrophilicity of these fibers was assessed by

contact angle measurements and 114 was attained on

silane treatment. SEM images showed the increase in

surface roughness and fibrillation on different fiber

treatments (Ramamoorthy et al. 2014). The treated

cellulose fibers were used to reinforce a synthesized

thermoset resin, and the obtained composites were

characterized in this study.

Tensile, flexural and impact properties were deter-

mined to see the mechanical performance of the

composites, in order find out the potential applications.

Tensile test

Table 1 shows tensile strength, Young’s modulus and

elongation for the different composite types. Nonwo-

ven composites were tested in machine direction and

the tensile strength of the composites was between 60

and 100 MPa which falls in line with previously

reported nonwoven fiber reinforced lactic acid based

thermoset composites (Esmaeili et al. 2014). Our

former study showed that soybean oil based thermoset

nonwoven fiber composites had tensile strength

between 70 and 100 MPa at 40–60 wt% fiber loading

(Ramamoorthy et al. 2012, 2013). Tensile modulus of

the composites was between 6 and 10 GPa while the

Table 1 Tensile and impact properties of the composites with various treated conditions

Sample

conditions

Tensile

strength (MPa)

Young’s

modulus (GPa)

Elongation (%) Impact

strength (kJ/m2)

Untreateda 80.86 ± 6.28 7.80 ± 1.41 2.22 ± 0.18 18.56 ± 1.60

Silane treated

6 wt% 82.41 ± 3.82 8.24 ± 0.83 1.68 ± 0.20 19.36 ± 2.51

8 wt% 85.67 ± 4.33 8.30 ± 2.04 1.59 ± 0.16 26.01 ± 3.51

10 wt% 91.43 ± 4.69 9.52 ± 1.89 1.24 ± 0.31 27.98 ± 2.46

Alkali treated

6 wt% 83.64 ± 5.57 8.10 ± 1.61 0.91 ± 0.20 23.78 ± 2.33

8 wt% 71.76 ± 5.93 7.83 ± 1.84 0.75 ± 0.18 18.93 ± 2.42

10 wt% 60.32 ± 7.82 6.90 ± 2.09 0.46 ± 0.10 16.92 ± 2.57

Thermally treatedb at 140 �C for 2 h (post-curing)

Untreated 79.15 ± 3.67 7.71 ± 1.35 2.11 ± 0.16 18.81 ± 1.34

Silane 10 wt% 96.29 ± 3.28 9.70 ± 1.64 1.16 ± 0.23 29.52 ± 3.80

Alkali 6 wt% 86.62 ± 5.10 8.56 ± 1.93 0.85 ± 0.19 24.47 ± 2.48

Standard deviation written after ±a No post curingb Thermal treatment

Cellulose

123

maximum elongation was about 2.1 %, which follow

the results from our earlier studies (Ramamoorthy

et al. 2012, 2013; Esmaeili et al. 2014). Further,

acoustic emissions could be used to analyze mi-

crostructural damage analysis (Bravo et al. 2015).

Tensile strength and modulus increased when the

fibers were treated with silane. Similar improvements

on silane treatments have been reported before (Faruk

et al. 2012; John and Anandjiwala 2008; Mohanty

et al. 2001). The changes were primarily due to the

amino functional silane, which not only alter the

hydrophilicity of the cellulose fibers but also react

with the resin in the interface. According to the

chemical bonding theory, the bifunctional silane

molecules act as a link between the cellulose fiber

and the resin; by forming a bond with fiber surface

through a siloxane bridge while its organofunctional

group bonds to the resin which gives molecular

continuity across the interface region of the composite.

Even at high silane concentration (10 wt%), the tensile

strength and the modulus of the composites in this

study did not reach 100 MPa and 10 GPa respectively.

This could be due to saturation of the interacting bonds

(fiber–Si–resin) and the textile architecture of the

reinforcement. As expected, the tensile elongation of

the composites decreased when the fibers were treated

with silane as the interface between fiber and matrix

was improved.

Tensile properties of alkali treated fiber reinforced

composites were dependent on concentration of the

alkali treatment; low concentration slightly increased

the strength and modulus of the composites whereas

high concentration decreased the tensile properties.

This marginal increase cannot be discussed due to high

standard deviation but the decrease was evident with

33 % lower tensile strength for highest alkali concen-

tration (10 wt%). This follows our preceding research

on alkali treatment on fibers where severe fibrillation

of fibers occurred at high concentration treatment

(Ramamoorthy et al. 2014). At low concentration, the

alkali treatment disrupts the hydrogen bonding in the

cellulose network and increases the surface roughness,

and good adhesion characteristics were expected when

reinforced in the matrix. The elongation of the

composites decreased with increasing alkali concen-

tration due to increased fiber splitting/fibrillation. A

small increase of surface roughness, pore formation

and fibrillation in fibers due to the alkali treatment

could improve the composite properties as the pores

on the fiber surface allow for good interlocking with

the matrix, whereas, high degree of pore formation

resulted in failure of the fibers.

Post curing of the composites was done because a

small exothermic peak was seen for the cured compos-

ites indicating incomplete curing, as discussed in the

DSC section. No exothermic peak was seen for post

cured samples. It has been reported that the impact and

the hardness properties of epoxy resins reinforced with

natural fibers were improved on post curing (Srinivasa

andBharath 2011). It is well-known that thorough cured

thermoset composites have longer lifetime than for

poorly cured composites. Though the increase in the

tensile properties of the post cured composites in this

work was not significant due to high standard deviation,

it is important to fully cure the thermoset resin to obtain

stable properties in structural composites. Partly cured

resins could result in high strain at rupture due to high

mobility of the polymer chains; giving a more ductile

behavior than that of completely cured ones.

Impact test

Charpy impact strength (Table 1) shows the energy

absorbed by the specimen during fracture. Impact

strength of the specimens was between 16 and 30 kJ/

m2. Low impact resistance of the produced composites

was expected due to alignment of the fibers in the

nonwoven fabrics. This is comparable to our earlier

study where nonwoven jute fiber was used as

reinforcement in soybean oil based thermoset resin

(Ramamoorthy et al. 2012). Similar results were

obtained when using cellulose fibers in a thermoplastic

polymer as reinforcement (Bledzki et al. 2009;

Oksman et al. 2003), and when using a lactic acid

based thermoset resin, which was impregnated in a

nonwoven natural fiber mat (Akesson et al. 2009). Our

prior studies showed that the impact strength increased

on changing the alignment of fibers (Ramamoorthy

et al. 2012; Esmaeili et al. 2014).

Silane treatment and low concentration alkali

treatment increased the impact strength whereas high

concentration alkali treatment decreased the impact

resistance. It is not possible to draw any conclusions,

as the change is small and not statistically significant.

Post-curing of the composites slightly increased the

impact strength as explained in previous section

(tensile properties). Impact strength followed a similar

trend to that of tensile strength and modulus.

Cellulose

123

Flexural test

Similarly to the tensile properties, the flexural prop-

erties of the composites, shown in Table 2, were

affected by fiber surface treatments. The flexural

strength and modulus of untreated fiber reinforced

composites were 116 MPa and 5 GPa respectively

which can be compared to the results from previous

studies (Ramamoorthy et al. 2013; Esmaeili et al.

2014). On silane treatment, there was insignificant

increase in flexural strength but the modulus increased

by 24 %. Meanwhile, the flexural strength and the

modulus fell drastically on alkali treatments. The

flexural strength and modulus dwindled 54 and 40 %

on highest alkali concentration. The reason for the

deterioration of the mechanical properties is discussed

in the previous sections. Post curing of the composites

increased the flexural strength and the modulus

marginally.

The data from the mechanical testing shows that the

properties of the composites could be improved on

appropriate chemical treatments such as silane and

alkali with suitable treatment conditions. The com-

posite could fail indeterminately due to poor interfa-

cial adhesion between fiber and matrix as the load

transfer amid the components is then ineffective. As a

result of poor interfacial adhesion, crack propagation

could be initiated at low stress due to brittleness of the

matrix. Therefore, it is necessary to improve the

interphase for good composite structures. The results

from this paper and the preceding paper show that the

mechanical properties are improved on surface treat-

ment due to enhanced interfacial adhesion between

fiber and matrix (Ramamoorthy et al. 2014).

Density and porosity measurements

Good resin penetration in the fabrics and the fiber

structures reduce the composite porosity. Furthermore,

the composite manufacturing process also influences

the porosity by trapping or releasing the air during

molding. Pores, which are air filled cavities, are

difficult to avoid in composites production due to the

mixing of sometimes high viscosity resins, and due to

dense reinforcements. Table 3 shows a summary of

the density and porosity measurements of the com-

posites. Fiber volume fraction in the composites was

maintained between 42 and 45 vol% and the porosity

volume was between 1 and 7 vol%. Increase in fiber

volume fraction could increase the tensile strength and

modulus whereas the pores develop due to improper

reinforcement wetting. These porous regions could

initiate the delamination of the layers resulting in

composite failure. A study based on natural fiber

composites showed that the porosity increased at high

fiber volume fraction (Madsen and Lilholt 2003).

Due to these reasons, a closely packed nonwoven

structure was chosen as reinforcement for a compos-

ite with a fiber volume percent between 42 and 45 %.

The pore volume was 6.4 % when the untreated

fabric was used as reinforcement and falls in line

with previous study on porosity prediction resulted in

porosity between 4 and 8 % (Madsen and Lilholt

2003). The high pore volume could be attributed to

the poor fiber–matrix interface. The treated fabrics

resulted in lower porosity of the composites; porosity

between 1.3 and 1.9 vol% on silane treatment while it

was between 3.4 and 4.3 vol% on alkali treatment,

the reason could be better interaction between fiber

and matrix and improved adhesion between fiber

surface and resin (introducing new moieties on silane

treatment and increase surface roughness on alkali

treatment). Lower porosity due to silane treatment

resulted in stronger composites; as higher tensile

properties were achieved, see Table 1. The decrease

in the porosity could also affect the water absorption

of the composites and this was explored later in this

paper.

Table 2 Flexural

properties of the composites

with various treated

conditions

Standard deviation written

after ±

Sample conditions Flexural strength (MPa) Flexural modulus (GPa)

Untreateda 115.9 ± 16.5 5.0 ± .1.3

Silane treated

6 wt% 121.9 ± 21.1 6.2 ± 1.0

Alkali treated

8 wt% 90.4 ± 18.3 3.2 ± 1.8

10 wt% 52.9 ± 17.4 3.0 ± 1.0

Cellulose

123

Dynamic mechanical thermal analysis (DMTA)

DMTA results showed the viscoelastic properties of

the cured resin and the composites. Storage modulus

of the cured resin was 3521 MPa at 30 �C which

corresponds to the polymer chain packing density in

the glassy state (Bakare et al. 2014). High cross-

linking density restricts the movement of the chain

segments. There was drop in storage modulus between

40 and 85 �C which is due to free movement of the

polymeric chain, also termed as rubbery plateau

region. Loss modulus of the cured resin was

358 MPa at 30 �C and the glass transition from tan dcurve peak was nearly 90 �C for the cured resin.

Figure 3a shows storage modulus of cured resin and

composites with and without surface treatments.

Storage modulus increased to 4206 MPa at 30 �C on

addition of fibers and this is due to the increase in

stiffness of composites compared to the cured resin.

The modulus increased to 4704 MPa at 30 �C on

introducing silane treated fibers (6 wt% conc) as

reinforcement due to good adhesion between fiber and

matrix; while it fell sharply on harsh alkali treatment

(10 wt% conc). These results are in a good agreement

with the mechanical properties, see Tables 1 and 2.

Figure 3b presents the tan delta curves of above

mentioned resin and composites; these curves indicate

the damping properties. The tan delta intensity of the

resin was high which demonstrates a balance between

elastic and viscous behavior of the resin (good

damping property) while the intensity of the tan delta

peak was limited when introducing the reinforcement,

as expected. The glass transition temperature of the

composites was over 100 �C; higher than that of the

neat resin. Tg increase on addition of cellulosic fibers

was due to restricted mobility of the polymer chains in

the interphase and was observed previously by several

authors (Akesson et al. 2011; Esmaeili et al. 2014).

Differential scanning calorimetry and thermogravi-

metric analysis results were investigated to study the

resin curing and thermal resistance. Thermal tests

were performed to determine exothermic heat release

from resin, glass transition temperature, storage

modulus, loss modulus and thermal degradation.

Differential scanning calorimetry (DSC)

DSC analysis showed the crosslinking reaction effi-

ciency of the resin at the used curing conditions and

the requirement of post-curing of the composites. No

exothermic peak was seen for the cured resin, which

indicates complete curing of the resin whereas the

uncured resin showed that 227 J/g exothermic heat

evolved, see Table 4. This high exothermic heat

corresponds to a high resin reactivity and short chain

length. The glass transition of the cured resin was

86 �C and this was lower than Tg (90 �C) obtainedfrom DMTA. This is mainly due to DMTA sensitivity

to the glass transition.

The produced composites were characterized to

monitor the resin cure, and it was found that the

composites needed a post-curing as a small exother-

mic heat (13 J/g) peak appeared. This exothermic heat

peak disappeared after the post-curing which confirms

Table 3 Density and porosity of the composites with various treated conditions

Sample

conditions

Density

(g cm-3)

Fiber volume

fraction (%)

Matrix volume

fraction (%)

Porosity volume

fraction (%)

Untreated 1.28 ± 0.08 42.22 ± 2.7 51.38 ± 3.2 6.40 ± 0.05

Silane treated

6 wt% 1.35 ± 0.02 44.49 ± 0.6 54.15 ± 0.8 1.36 ± 0.01

8 wt% 1.35 ± 0.04 44.55 ± 1.3 54.22 ± 1.6 1.23 ± 0.03

10 wt% 1.34 ± 0.09 44.25 ± 2.9 53.85 ± 3.6 1.90 ± 0.07

Alkali treated

6 wt% 1.32 ± 0.04 43.43 ± 1.4 52.85 ± 1.7 3.72 ± 0.03

8 wt% 1.32 ± 0.01 43.56 ± 0.5 53.01 ± 0.6 3.43 ± 0.01

10 wt% 1.31 ± 0.04 43.18 ± 1.3 52.55 ± 1.6 4.27 ± 0.03

Standard deviation written after ±

Cellulose

123

the complete curing of the composites. The tensile

strength (Table 1), did not increase significantly on

post-curing but it is necessary to have thorough curing

of the resin in the composite to minimize the failure

nodes in structural applications (Karbhari 2007). The

ductility of the partially cured resins is good as the

stain at break is high due to mobility of the polymer

chains; consequently it reduces the tensile properties

of the composites.

Thermogravimetric analysis (TGA)

Thermal decomposition resistance of the resin and the

composites were analyzed by studying the gravimetric

weight loss with the function of temperature. The resin

was relatively stable up to 240 �C and the maximum

degradation occurred after 375 �C. At 259 �C, the

weight loss was about 10 % whereas 90 %weight loss

and second derivative peak were at 442 and 428.5 �C

Fig. 3 a Storage modulus

of cured resin and

composites with and without

reinforcement surface

treatments (silane and

alkali). b Tan delta of cured

resin and composites with

and without reinforcement

surface treatments (silane

and alkali)

Cellulose

123

respectively. The weight loss of the composites varied

on composites due to the inclusion of cellulose fibers.

Initial ten percent weight loss of the composites was

mainly because of the amount of the moisture in the

composites, and slight decomposition of the resin. The

temperature at which 90 % of the material is lost was

increased from 442 �C in the resin to 595 �C in the

composites and the treatments on reinforcements did

not vary the results significantly. The increase in

degradation temperature for the composites compared

to neat resin was mainly due to the residual ash

obtained from high amount reinforcement (50 %) in

the composites. Fiber surface treatment did not affect

the thermal degradation, as all components of the

composites degrade at high temperature. The second

derivative curves showed two peaks for the compos-

ites due to degradation of reinforcement and matrix at

different temperatures. The second derivative peak for

the matrix was between 417 and 424 �C; and it

reduced slightly in the composites whereas the peak

for neat reinforcement was between 268 and 310 �C.This change in derivative peak temperature for the

reinforcements is interesting as it shows the effect of

surface treatments. Rapid cellulose degradation occurs

at about 300 �C and this could shift on surface

treatments as shown in Table 5.

Contact angle and surface energy

Polarity of the fibers and the matrix are responsible for

the hydrophilicity of the composites. High contact

angle value of the composites could be due to the lack

of polar surface groups or that these groups have non-

accessible locations. When a liquid comes in a contact

with a solid surface, there is an interaction of the polar

and dispersive parts of both solid and liquid phase at

the interface. The surface tension at the interface is

higher than the total surface tension of the solid and

liquid phase, which is due to interactions at the

interface (Ramamoorthy et al. 2014). Table 6 shows

the surface energy components of the probe liquids

used in the wetting experiments. The probe liquids

were used to measure the contact angle (h) and the

acid–base surface energy. The contact angle is a

function of the solid’s surface energy and the liquid’s

surface tension. The contact angle below 90� repre-

sents the wetting of the solid phase whereas the angle

below 90� represents the non-wettability.The values of the contact angle with standard

deviation of all composites for four probe liquids are

given in Table 7. The contact angle of the composites

increased when the reinforcements were treated with

silane or alkali except for the 6 wt% alkali treatment.

These results were coherent with previous studies

when natural and regenerated cellulose fibers were

surface treated with silane and alkali (Ramamoorthy

et al. 2014; Park et al. 2006). The high contact angle of

about 59� and 62� was achieved on treating the fibers

with silane and alkali treatments respectively. There

could be several reasons for the change in the contact

angle such as silane molecules covering hydrophilic

fiber surface, good penetration of matrix into alkali

treated fiber pores, good interlocking of matrix with

surface treated fibers and good interfacial adhesion

between fiber and matrix. The composites are still

Table 4 DSC characterization results of the resin and the

composites

Exothermic

heat (J/g)

Tg (�C)

Resin

Uncured 227.4 –

Cured 0 85.8

Composite

Before post-curing 13.0 81.8

After post-curing 0 81.7

Table 5 Thermal stability results from TGA characterization

TGA

Temperature

at 10 wt%

Loss (�C)

Temperature

at 90 wt%

Loss (�C)

Second

Derivative

Peaks (�C)

Resin

Cured 258.9 442.1 428.5

Composite

Untreated 254.8 595.5 308.6; 417.6

Silane treated

6 wt% 262.7 595.3 309.2; 417.3

8 wt% 258.7 595.4 302.6; 424.2

10 wt% 252.6 595.4 297.2; 423.1

Alkali treated

6 wt% 261.1 593.2 297.4; 420.0

8 wt% 261.4 595.3 268.7; 424.8

10 wt% 262.0 595.4 285.3; 424.7

Cellulose

123

hydrophilic as the contact angle of the composites is

lower than 90�; this outcome supports the results from

water absorption. Nevertheless, the change in contact

angle clearly indicates the change in hydrophilicity of

the composites on surface treatments with silane and

alkali.

The total surface free energy, ctot, is the sum of

dispersive (cd) and the polar components (cp). Thedispersive and polar surface energies are measured by

Owens–Wendt equation,

cL 1þ cos hð Þ ¼ 2 cdScdL� �1=2þ2 cpScpL

� �1=2 ð8Þ

where cL, cdL (dispersive) and cpL (polar) are known

surface energies of test liquids and cdS and cpS are

surface energies of the solid which are calculated from

the measured contact angles.

The acid–base component (hydrogen bonding),

cAB, includes electron acceptor (c?) and electron

donor (c-) components. It is calculated by

cAB ¼ 2 cþc�ð Þ1=2 ð9Þ

Table 8 shows the surface energy components of

the composites. The surface energy decreased on

silane and alkali treatments; except when the fiber was

treated with 6 wt% concentration alkali. The decrease

in the surface energy indicates the increase in

hydrophobicity. The results follow the outcome from

the contact angles. The decrease of the basic compo-

nent, c-, of the surface energy indicates the reductionof hydroxyl groups. The silane treatment reduces the

c- as the hydroxyl groups were covered by silane

Table 6 Constant surface energy components of probe liquids

Solvent (heavy phase) ctot (mN/m) cd (mN/m) cp (mN/m) c? (mN/m) c- (mN/m) cAB (mN/m)

Formamide 58 39 19 2.28 39.6 19

Ethylene–glycol 48 29 19 3 30.1 19

Water 72.8 21.8 51 25.5 25.5 51

Hexadecane 27.47 27.47 0 0 0 –

Light phase—air

Table 7 Contact angles

with standard deviation of

the composites using optical

tensiometer

Conditions Contact angle (h)

Water Formamide Ethylene–glycol Hexadecane

Untreated 42.4 ± 4.7 18.3 ± 3.6 33.3 ± 2.1 10.0 ± 1.9

Silane treated

6 wt% 45.3 ± 4.5 21.8 ± 2.9 25.1 ± 3.1 18.8 ± 1.1

8 wt% 52.1 ± 3.6 24.0 ± 2.0 33.0 ± 2.8 20.8 ± 5.8

10 wt% 58.6 ± 5.3 36.2 ± 6.8 37.0 ± 4.2 26.7 ± 3.0

Alkali treated

6 wt% 36.8 ± 3.4 20.7 ± 3.6 27.7 ± 2.0 18.6 ± 2.3

8 wt% 61.8 ± 3.5 36.8 ± 4.3 37.0 ± 4.2 21.4 ± 1.7

10 wt% 62.2 ± 1.1 41.3 ± 4.0 42.7 ± 4.0 26.9 ± 3.6

Table 8 Acid–base surface energy components of the

composites

Conditions ctot cd cp sqrt(c?) sqrt(c-)

Untreated 54.3 9.4 44.9 4.0 5.5

Silane treated

6 wt% 53.2 9.0 44.2 4.3 5.1

8 wt% 47.3 9.9 37.4 4.2 4.5

10 wt% 41.8 10.1 31.7 3.9 4.1

Alkali treated

6 wt% 59.2 8.0 51.2 4.2 6.0

8 wt% 39.1 11.6 27.5 3.8 3.6

10 wt% 38.7 10.8 27.9 3.5 3.9

The values are calculated from the mean values in Table 7

Cellulose

123

coupling agents. The surface energy and the contact

angle of the neat fibers were investigated our previous

study (Ramamoorthy et al. 2014). The results of the

fiber and the composite fall in line with each other;

with these results further investigations on work of

adhesion could be done (Park et al. 2006).

Water absorption

Water absorption of the composites was followed

for 10 days and it was found that the absorption was

more pronounced for the first 5 days. The absorption

was minor after first 5 days as observed previously

with cellulose fiber composites (Ramamoorthy et al.

2012). The absorption of the composites was

primarily influenced by the cellulose reinforcement

(Esmaeili et al. 2014). It could also be influenced by

diffusion of water molecules between the polymer

chain, into the pores and into the fiber–matrix

interface. So, it is necessary to use less hydrophilic

fibers and a resin with higher cross-linking density.

The silane and alkali treatments on reinforcements,

and their concentrations affected the water absorp-

tion of the composites, see Fig. 4. The silane

treatment, 10 wt% concentration, reduced the water

absorption of the composites from 33 to 23 wt%

while the reduction was not conspicuous on alkali

treatment for same concentration. The reduction

could be due to two reasons, namely the chemical

modification in the fiber and the interfacial changes

in the composites. These results were in line with

porosity measurements and mechanical properties.

The mechanical properties, strength and modulus, of

the composites deteriorated after absorption and this

was in an agreement with former study (Ramamoor-

thy et al. 2012). The percentage decrease in tensile

properties is tabulated in Table 9. The tensile

strength and modulus fell 79 and 66 % respectively

on alkali treatment while this was lower on silane

treatment, 51 and 58 % correspondingly.

Table 9 Water absorption

of the composites with

various treated conditions

The percentage reduction is

calculated from the mean

strength and modulus

before and after water

absorption

Sample conditions % Reduction in tensile properties after water absorption

Tensile strength (%) Young’s modulus (%)

Silane treated

6 wt% 34.4 49.4

8 wt% 45.8 59.9

10 wt% 51.0 58.2

Alkali treated

6 wt% 64.2 56.3

8 wt% 65.9 59.7

10 wt% 79.4 65.8

Fig. 4 Water absorption

(wt%) of the composites

after 10 days immersion in

water

Cellulose

123

Microscopy

Figure 5 shows the SEM micrographs of tensile frac-

tured surface before and after water absorption. Un-

treated fiber reinforced composite before water

absorption (a) had pores at the interface and fiber pull

out was noticed whereas silane treated composite

(b) was well embedded in the matrix and had smaller

fiber pull out. This ratifies the increase in tensile strength

and modulus, and decrease in elongation on silane

treatment (Table 1). Alkali treated composite (c) had

pores and fiber damage which confirms the decrease in

tensile properties on alkali treatment. The fiber damage

resulted in low elongation due to failure of fiber and

composites at low strain. These deductions go together

with the results from porosity and water absorption.

Fig. 5 SEM images of tensile fractured composites; before water absorption a untreated, b silane 10 wt%, c alkali 10 wt%; after water

absorption d untreated, e silane 10 wt% f alkali 10 wt%

Fig. 6 Optical microscopy images of tensile fractured composites after water absorption a silane 10 wt% b alkali 10 wt% treated fiber

reinforced composites

Cellulose

123

Fig. 7 Resonance analysis

using Audacity and Scilab

a sound data, amplitude vs

time, obtained using

Audacity b frequency data

converted from sound data,

amplitude intensity versus

frequency c graph plotted,

frequency against different

length

Cellulose

123

Water absorbed specimens behaved differently

during tensile test, Fig. 5d–f. There was long fiber

pull out for all specimens after water absorption;

which reduced the tensile strength and modulus

(Table 9), and increased the elongation.

Optical microscopy images were taken to see the

behavior of layers on water absorbed tensile fractured

specimens. The delamination of layers was noticed on

treated water absorbed specimens. The reduction of

tensile properties (Table 9) was partly due to de-

lamination caused by water absorption at interface.

This also confirms the water absorption at interface

(Fig. 6).

Resonance analysis

The modulus calculated using this non-destructive

method was of insignificant difference to the modulus

obtained from destructive tensile test, which conclude

that both approaches lead to a correct estimate of

Young’s modulus. The acoustic signals of the vibrat-

ing composite beam were detected as a function of

amplitude to time, as shown in Fig. 7a. Harmonic

oscillations on displacing the beam (mechanical

vibrations) were converted into an electrical signal

using piezoelectric transducer before the signals were

recorded by the sound card in Audacity. The time data

was converted into frequency by a fast Fourier

transform (FFT) using Scilab, as shown in Fig. 7b.

The script used to execute the FFT is shown in

characterization section. Tests were repeated with the

different beam fastening lengths and the obtained

graph was plotted as show in Fig. 7c. The slope

obtained from the Fig. 7d and the modulus was

calculated from Eq. 4.

The modulus of the 8 and 10 wt% concentration

alkali treated fiber composites was 7.2 and 6.7 GPa

respectively on non-destructive analysis, whereas the

same composites yielded 7.8 and 6.9 GPa on destruc-

tive tensile testing. Likewise, the modulus of the silane

treated fiber composites was in agreement with tensile

testing. Similar analysis was performed on concrete

slabs as a part of French national research project

(Villain et al. 2011). Resonance analysis can also be

used to determine the fiber debonding, delamination,

cracking and micro-failure mechanism inside the

composite structure (Park et al. 2006). It is also

known that torsional vibrations yield shear modulus of

the composites.

Thermo-physical test

Table 10 shows the bulk thermo-physical properties of

the composites. Thermal conductivity, thermal diffu-

sivity and specific heat were measured simultaneously

at constant temperature as these properties were

largely affected by temperature. There was an increase

in the conductivity and diffusivity of the composites

on fiber surface treatment, whereas the specific heat of

the composites decreased on treatments. This was in

agreement with our previous study on fibers where the

Cp was reduced from 1.66 and 1.25 to 1.19 and 0.99 J/

g C on alkali and silane treatments respectively

(Ramamoorthy et al. 2014). Thermal conductivity of

natural fibers and thermoset acrylic resins has been

studied previously by Behzad and Sain (2007).

Thermal conductivity of the resin (0.43 W/m K) was

lower than that of the fiber (1.48 W/m K). The

composite produced in our study possess low thermal

conductivity, 0.363 W/m K, when the fabric was

reinforced without any treatment. The low conduc-

tivity was mainly due to the fact that 50 % of the

composites were thermoset matrix. On silane and

alkali treatment, the conductivity was 0.4 and 0.38 W/

m K respectively. These results were in line with oil

palm fiber reinforced phenolformaldehyde composites

where the conductivity of the composites (0.29 W/

m K) was increased when the reinforcing fiber was

treated with silane (0.46 W/m K) and alkali (0.42 W/

m K) (Agrawal et al. 2000). The increase on silane

treatment could be due to silane molecules available

on fiber surface, which make the fiber less hydrophilic

and increases the adhesion between the treated fiber

and thermoset matrix. It can also be due to the increase

in the polarity due to the presence of Si–OH group.

Alkali treatment also increased the thermal conduc-

tivity, which could be a result of improved adhesion

between fiber and matrix as the OH groups react with

NaOH and gives rough surface topography. The

Table 10 Thermal conductivity, diffusivity and specific heat

measured using TPS 2500 hot disk

Composite Conductivity

(W/m K)

Diffusivity

(mm2/s)

Specific heat

(MJ/m2K)

Untreated 0.363 0.284 1.29

Silane 8 wt% 0.400 0.333 1.20

Alkali 8 wt% 0.379 0.296 1.28

Cellulose

123

change in the fiber’s surface roughness and increase in

diameter was noticed using microscopy images in our

preceding study (Ramamoorthy et al. 2014). Similar

fiber behavior on surface treatment was noticed in this

study. The pores created on alkali treatment allow

good interlocking with the matrix to increase the

thermal conductivity.

Thermal diffusivity of the composites also in-

creased on fiber treatments and this was in good

agreement with oil palm fiber reinforced phenol-

formaldehyde composites (Agrawal et al. 2000). The

specific heat capacity of the composites decreased on

fiber treatment, which shows that lower energy was

required to heat the composites after treatment. Heat

capacity of the treated fibers was discussed in our

preceding study (Ramamoorthy et al. 2014).

Finite element analysis

From our preceding study, it was found that the Cp of

the regenerated cellulose fiber was reduced from 1.66

and 1.25 to 1.19 and 0.99 J/g C on alkali and silane

treatments respectively (Ramamoorthy et al. 2014).

The specific heat increased as the temperature rose

from 30 to 50 �C. Hemp fiber showed similar increase

in Cp from 2.2 J/g K to almost 3.4 J/g K when the

temperature rose from 20 to 100 �C (Behzad and Sain

2007). Similarly, the specific heat of the cured resin

increased with temperature and could be fitted in

second order polynomial as explained by Behzad and

Sain (2007).

Composite was modeled in COMSOL 4.3b with

same amount of fiber and resin as in the experiment.

This simulation was carried out on some assumptions;

composite laminate is free of voids, heat loss due to

radiation and convection is neglected, fibers are

uniform (shape and size), aligned uni-directionally

and thermal contact resistance between fiber and

matrix interface is negligible. The study was done with

heat transfer module in time dependent model by

employing the experimental values of thermal con-

ductivity and specific heat (transverse direction–

across the fiber). The temperature variation with time

at different locations was obtained. It was noted that

the predicted temperature data was in agreement with

the measured data with slight disparity, and similar

results were obtained with hemp fiber reinforced

composites (Behzad and Sain 2007). Thermal con-

ductivity of the composites could be also predicted by

E–S model as explained by Behzad and Sain (2007).

As expected, the results changed when the model was

studied along the fiber (in-plane direction) and when

the amount of fiber increased/decreased in the

composites.

However, the numerical results in this paper are

limited as the moisture uptake resulting in property

degradation and the large variation of the fiber

properties after treatment was not considered. Detailed

numerical study of the hemp fiber composite’s mois-

ture uptake and mechanical properties was done by

Toubal et al. (2015) which will be considered in our

future research.

Conclusions

Detailed study of biocomposites produced from

regenerated cellulose fibers and lab synthesized resin

was discussed in order to improve the interface and

composite properties. Regenerated cellulose fibers

were treated with silane and alkali and their properties

were thoroughly discussed in our preceding work

(Ramamoorthy et al. 2014) and the resin was charac-

terized in our former study (Bakare et al. 2014).

Biocomposites were manufactured by reinforcing

untreated and treated regenerated cellulose fibers in

novel lab synthesized bioresin (78 % bio content)

from glycerol and lactic acid by direct condensation

and functionalization. Tensile, flexural and impact

tests were used to evaluate the mechanical properties

of the composites whereas DMTA was used to

evaluate viscoelastic properties. The composites had

good mechanical properties and the properties were

improved by some treatments. Highest tensile strength

(91 MPa) and modulus (9.5 GPa) were obtained when

the fiber was treated with silane. The improvement on

silane treatment was due to molecular continuity

formed in the interphase of the composite as silane

molecules can form siloxane bridges between the

cellulose fiber and the resin. Alkali treatments also

altered the properties which is mainly due to increase

in surface roughness and hydrogen bonding disrup-

tion. Higher strength could be obtained by changing

the fabric architecture as discussed in our previous

study (Esmaeili et al. 2014). Hybridization could also

improve the properties as discussed in former study

(Ramamoorthy et al. 2012). The results from the

impact and the flexural tests fall in line with the tensile

Cellulose

123

test results; highest impact strength (28 kJ/m2),

flexural modulus (122 MPa) and flexural strength

(6.2 GPa) were obtained when the fibers were treated

with silane. Highest viscoelastic properties were

achieved for the composites reinforced with silane

treated fibers and follow the results from the me-

chanical tests. Glass transition temperature of the

composites was measured from DMTA tan delta

curves and was over 100 �C. Archimedes principle

was used to determine density and porosity measure-

ments. Porosity volume of the composites reduced

from 6.4 to 1.2 and 3.4 % on silane and alkali

treatment respectively. This was due to interface

improvement as the fibers were well embedded in the

matrix. Thermal properties were assessed by DSC and

TGA. The curing of the resin was followed using DSC

and the composites were subjected to post-curing

based on DSC results. The glass transition temperature

of synthesized resin was higher than thermoplastic

PLA. Glass transition fromDSCwas lower than that of

DMTA due to the sensitivity of DMTA. Thermal

degradation was followed using TGA and the change

in the second derivative peak showed that the treat-

ments altered the stability of the reinforcement. The

hydrophilicity of the composites was evaluated using

contact angle, surface energy and water absorption.

Fiber treatment changed the hydrophilicity of the

composites and this was due to changes in the fibers

and interface. Water absorption decreased the tensile

properties of the composites as expected. Morpho-

logical analysis performed using microscopy images.

These images show the good interface between silane

treated fiber and matrix whereas the untreated fiber

composites had pores at interface. This indicated

better adhesion between fiber and resin on silane

treatment. The fiber damage was noticed on high alkali

concentration and long fiber pull out was noticed on

water absorbed specimens. These results get along

with the tensile properties. Thermal conductivity,

diffusivity and specific heat were measured for the

composites, and it was found that the treatment

improved the conductivity and diffusivity as anticipat-

ed. Non-destructive method, resonance analysis, was

used to measure modulus of the composites by

converting the acoustic signals of the vibrating

composite beam to electrical signals with piezoelectric

transducer. The modulus was similar to the results

from the destructive tensile testing. Finite element

analysis was used to predict the thermal behavior of

the composites. The temperature variation obtained

from simulation was comparable to the experimental

results regardless of some assumptions. However, the

modeling in this paper was limited as the moisture

uptake resulting in property degradation and the large

variation of the fiber properties after treatment was not

considered.

The results from this paper show the potential to use

the previously reported novel bioresin and helps in

choosing a surface treatment for cellulose fibers

intended for composites. This paper also shows the

potential of regenerated cellulose fibers in structural

composites. The properties of these composites could

be further enhanced by changing the textile architec-

ture and physical characteristics of the reinforcements

as discussed in our previous study, and can be used in

various non-structural applications (Esmaeili et al.

2014). The changes in the composite properties were

due to the changes in fiber surface and interface.

Previous research on surface modified natural fibers

showed better performance in composite production

and their properties due to the change in hydrophilicity

of the fiber and fiber–matrix adhesion. These com-

posites were subsequently used in composite applica-

tions. In our upcoming project, the stability of the

cellulose composites on water absorption will be

focused. On further improvement, regenerated cellu-

lose fiber reinforced composites could also be used in

several composite applications.

Acknowledgments This research was funded by research

foundation AForsk, Sweden. Authors would like to thank Adib

Kalantar for assisting in thermal conductivity measurements.

References

Adekunle K, Patzelt C, Kalantar A, Skrifvars M (2011) Me-

chanical and viscoelastic properties of soybean oil ther-

moset reinforced with jute fabrics and carded lyocell fiber.

J Appl Sci 122:2855–2863

Adusumali RB, Reifferscheid M, Weber H, Roeder T, Sixta H,

Gindl W (2006) Mechanical properties of regenerated cel-

lulose fibres for composites. Macromol Symp 244:119–125

Agrawal R, Saxena NS, Sreekala MS, Thomas S (2000) Effect

of treatment on the thermal conductivity and thermal dif-

fusivity of oil-palm-fiber-reinforced phenolformaldehyde

composites. J Polym Sci Part B Polym Phys 38:916–921

Akesson D, Skrifvars M, Seppala J, Turunen M, Martinelli A,

Matic A (2009) Processing of structural composites from

biobased thermoset resins and natural fibres by compres-

sion moulding. J Biobased Mater Bioeneg 3:215–225

Cellulose

123

Akesson D, Skrifvars M, Seppala J, Turunen M, Martinelli A,

Matic A (2010) Synthesis and characterization of a lactic

acid-based thermoset resin suitable for structural compos-

ites and coatings. J Appl Sci 115:480–486

Akesson D, Skrifvars M, Seppala J, Turunen M (2011) Ther-

moset lactic acid-based resin as a matrix for flax fibers.

J Appl Sci 119:3004–3009

Bakare FO, Skrifvars M, Akesson D, Wang Y, Afshar SJ, Es-

maeili N (2014) Synthesis and characterization of bio-

based thermosetting resins from lactic acid and glycerol.

J Appl Polym Sci. doi:10.1002/app.40488

Behzad T, Sain M (2007) Measurement and prediction of

thermal conductivity of hemp fibre reinforced composites.

Polym Eng Sci 47:977–983

Bledzki AK, Gassan J (1999) Composites reinforced with cel-

lulose based fibres. Progr Polym Sci 24:221–274

Bledzki AK, Reihmane S, Gassan J (1996) Properties and

modification methods for vegetable fibers for natural fiber

composites. J Appl Polym Sci 59:1329–1336

Bledzki AK, Jaszkiewicz A, Scherzer D (2009) Mechanical

properties of PLA composites with man-made cellulose

and abaca fibres. Compos Part A 40:404–412

Bravo A, Toubal L, Koffi D, Erchiqui F (2015) Development of

novel green and biocomposite materials: tensile and flex-

ural properties and damage analysis using acoustic emis-

sion. Mater Design 66:16–28

Carrillo F, Colom X, Canavate X (2010) Properties of regen-

erated cellulose lyocell fiber-reinforced composites. J Re-

inf Plast Compos 29:359–371

Caulfield DF, Feng D, Prabawa S, Young RA, Sanadi AR (1999)

Interphase effects on the mechanical and physical aspects of

natural fiber composites. AngewMakromol Chem 272:57–64

Esmaeili N, Bakare FO, Skrifvars M, Afshar SJ, Akesson D

(2014) Mechanical properties for bio-based thermoset

composites made from lactic acid, glycerol and viscose

fibers. Cellulose. doi:10.1007/s10570-014-0500-3

Faruk O, Bledzki AK, Fink HP, Sain M (2012) Biocomposites

reinforced with natural fibers: 2000–2010. Progr Polym Sci

37:1552–1596

Fink HP, Ganster J, Lehmann A (2014) Progress in cellulose

shaping: 20 years industrial case studies at Fraunhofer

IAP. Cellulose 21:31–51

Ganster J, Fink HP (2006) Novel cellulose fibre reinforced

thermoplastic materials. Cellulose 13:271–280

Jaturapiree A, Manian AP, Bechtold T (2006) Sorption studies

on regenerated cellulose fibers in salt–alkali mixtures.

Cellulose 13:647–654

Jawaid M, Abdul Khalil HPS (2011) Cellulosic/synthetic fibre

reinforced polymer hybrid composites: a review. Carbohyr

Polym 86:1–18

John MJ, Anandjiwala RD (2008) Recent developments in

chemical modification and characterization and charac-

terization of natural fiber-reinforced composites. Polym

Compos 29:187–207

Johnson RK, Zink-Sharp A, Renneckar SH, Glasser WG (2008)

Mechanical properties of wetlaid lyocell and hybrid fiber-

reinforced composites with polypropylene. Compos Part A

Appl Sci Manuf 39:470–477

Karbhari VM (ed) (2007) Durability of composites for civil

structural applications. Woodhead Publishing, CRC Press,

Florida, p 61

Koronis G, Silva A, Fontul M (2013) Green composites: a re-

view of adequate materials for automotive applications.

Compos Part B 44:120–127

La Mantia FP, Morreale M (2011) Green composites: a brief

review. Compos Part A 42:579–588

Madsen B, Lilholt H (2003) Physical and mechanical properties

of unidirectional plant fibre composites—an evaluation of

the influence of porosity. Compos Sci Technol

62:1265–1272

Madsen B, Thygesen A, Lilholt H (2007) Plant fibre–porosity

and volumetric interaction. Compos Sci Technol

67:1584–1600

Mohanty AK, Misra M, Drzal LT (2001) Surface modifications

of natural fibers and performance of the resulting bio-

composites: an overview. Compos Interf 8:313–343

Oksman K, Skrifvars M, Selin JF (2003) Natural fibres as re-

inforcement in polylactic acid (PLA) composites. Compos

Sci Technol 63:1317–1324

Okubayashi S, Bechtold T (2005) Alkali uptake and swelling

behaviour of lyocell fiber and their effects on crosslinking

reaction. Cellulose 12:459–467

Park JM, Quang ST, Hwang BS, DeVries KL (2006) Interfacial

evaluation of modified jute and hemp fibers/polypropylene

(PP)-maleic anhydride polypropylene copolymers (PP-

MAPP) composites using micromechanical technique and

non-destructive acoustic emission. Compos Sci Technol

66:2686–2699

Puglia D, Biagiotti J, Kenny JM (2005) A review on natural

fibre-based composites—part II: application of natural re-

inforcements in composite materials for automotive in-

dustry. J Nat Fib 1:23–65

Ramamoorthy SK, Di Q, Adekunle K, Skrifvars M (2012) Effect

of water absorption on mechanical properties of soybean

oil thermosets reinforced with natural fibers. J Reinf Plast

Compos 31:1191–1200

Ramamoorthy SK, Kundu CK, Adekunle K, Bashir T, Skrifvars

M (2013) Properties of green composites with regenerated

cellulose fiber and soybean-based thermoset for technical

applications. J Reinf Plast Compos. doi:10.1177/

0731684413504325

Ramamoorthy SK, Skrifvars M, Rissanen M (2014) Effect of

alkali and silane surface treatments on regenerated cellu-

lose fibre type (Lyocell) intended for composites. Cellu-

lose. doi:10.1007/s10570-014-0526-6

Ramamoorthy SK, Skrifvars M, Persson A (2015) A review of

natural fibers used in biocomposites: plant, animal and

regenerated cellulose fibers. Polym Rev 55:107–162.

doi:10.1080/15583724.2014.971124

Saevey KC, Ghosh I, Davis RM, GlasserWG (2001) Continuous

cellulose fiber-reinforced cellulose ester composites I.

Manufacturing options. Cellulose 8:149–159

Seniha Guner F, Yagci Y, Tuncer Erciyes A (2006) Polymers

from triglyceride oils. Progr Polym Sci 31:633–670

Srinivasa CV, Bharath KN (2011) Impact and hardness prop-

erties of areca fiber–epoxy reinforced composites. J Mater

Environ Sci 2:351–356

Toubal L, Cuilliere JC, Bensalem K, Francois V, Gning PB

(2015) Hygrothermal effect on moisture kinetics and me-

chanical properties of hemp/polypropylene composite:

experimental and numerical studies. Polym Compos.

doi:10.1002/pc.23414

Cellulose

123

Villain G, Marrec LL, Rakotomanana L (2011) Determination

of bulk elastic moduli of various concretes by resonance

frequency analysis of slabs submitted to impact echo. Eur J

Environ Civ Eng 15:601–617

Woodings C (ed) (2001) Regenerated cellulose fibres. Wood-

head Publishing, CRC Press, Boston. ISBN 978-1-85573-

459-3

Cellulose

123